|
|
""" |
|
|
Maya-1-Voice VLLM Streaming Inference - Standalone Reference Implementation |
|
|
|
|
|
This is a complete, self-contained example for using Maya-1-Voice TTS model with VLLM and SNAC. |
|
|
Demonstrates streaming audio generation with sliding window approach for smooth playback. |
|
|
|
|
|
Requirements: |
|
|
pip install vllm transformers torch snac numpy |
|
|
|
|
|
Usage: |
|
|
python vllm_streaming_inference.py |
|
|
|
|
|
Author: Maya-1-Voice Team |
|
|
License: MIT |
|
|
""" |
|
|
|
|
|
import torch |
|
|
import numpy as np |
|
|
import asyncio |
|
|
from typing import List, Optional, AsyncGenerator |
|
|
from transformers import AutoTokenizer |
|
|
from vllm import AsyncLLMEngine, AsyncEngineArgs, SamplingParams |
|
|
from snac import SNAC |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
CODE_START_TOKEN_ID = 128257 |
|
|
CODE_END_TOKEN_ID = 128258 |
|
|
CODE_TOKEN_OFFSET = 128266 |
|
|
|
|
|
|
|
|
SNAC_MIN_ID = 128266 |
|
|
SNAC_MAX_ID = 156937 |
|
|
|
|
|
|
|
|
SNAC_MODEL_NAME = "hubertsiuzdak/snac_24khz" |
|
|
SNAC_SAMPLE_RATE = 24000 |
|
|
SNAC_TOKENS_PER_FRAME = 7 |
|
|
|
|
|
|
|
|
DEFAULT_TEMPERATURE = 0.4 |
|
|
DEFAULT_TOP_P = 0.9 |
|
|
DEFAULT_MAX_TOKENS = 2000 |
|
|
DEFAULT_MIN_TOKENS = 28 |
|
|
DEFAULT_REPETITION_PENALTY = 1.1 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
class SNACDecoder: |
|
|
""" |
|
|
Decodes SNAC tokens (7-token frames) to audio waveforms. |
|
|
|
|
|
The unpacking logic converts flat 7-token frames back to hierarchical |
|
|
3-level SNAC codes (matching the training preprocessing exactly). |
|
|
""" |
|
|
|
|
|
def __init__(self, device: str = "cuda"): |
|
|
"""Initialize SNAC decoder with 24kHz model.""" |
|
|
self.device = device |
|
|
print(f"๐ต Loading SNAC 24kHz model to {device}...") |
|
|
self.snac_model = SNAC.from_pretrained(SNAC_MODEL_NAME).eval().to(device) |
|
|
print(f"โ
SNAC decoder initialized") |
|
|
|
|
|
def unpack_snac_from_7(self, vocab_ids: List[int]) -> List[List[int]]: |
|
|
""" |
|
|
Unpack 7-token SNAC frames to 3 hierarchical levels. |
|
|
|
|
|
This is the EXACT INVERSE of training preprocessing. |
|
|
|
|
|
Frame structure (7 tokens per frame): |
|
|
[slot0, slot1, slot2, slot3, slot4, slot5, slot6] |
|
|
|
|
|
Unpacking to [L1, L2, L3]: |
|
|
- slot0 โ L1[i] (coarse: 1x rate) |
|
|
- slot1 โ L2[2*i] (medium: 2x rate, even) |
|
|
- slot2 โ L3[4*i+0] (fine: 4x rate) |
|
|
- slot3 โ L3[4*i+1] |
|
|
- slot4 โ L2[2*i+1] (medium: odd) |
|
|
- slot5 โ L3[4*i+2] |
|
|
- slot6 โ L3[4*i+3] |
|
|
|
|
|
Args: |
|
|
vocab_ids: List of SNAC token IDs (128266-156937), length divisible by 7 |
|
|
|
|
|
Returns: |
|
|
[L1, L2, L3] where L1=n, L2=2n, L3=4n elements |
|
|
""" |
|
|
|
|
|
if vocab_ids and vocab_ids[-1] == CODE_END_TOKEN_ID: |
|
|
vocab_ids = vocab_ids[:-1] |
|
|
|
|
|
|
|
|
frames = len(vocab_ids) // SNAC_TOKENS_PER_FRAME |
|
|
vocab_ids = vocab_ids[:frames * SNAC_TOKENS_PER_FRAME] |
|
|
|
|
|
if frames == 0: |
|
|
return [[], [], []] |
|
|
|
|
|
l1, l2, l3 = [], [], [] |
|
|
|
|
|
for i in range(frames): |
|
|
slots = vocab_ids[i*7:(i+1)*7] |
|
|
|
|
|
|
|
|
l1.append((slots[0] - CODE_TOKEN_OFFSET) % 4096) |
|
|
l2.extend([ |
|
|
(slots[1] - CODE_TOKEN_OFFSET) % 4096, |
|
|
(slots[4] - CODE_TOKEN_OFFSET) % 4096, |
|
|
]) |
|
|
l3.extend([ |
|
|
(slots[2] - CODE_TOKEN_OFFSET) % 4096, |
|
|
(slots[3] - CODE_TOKEN_OFFSET) % 4096, |
|
|
(slots[5] - CODE_TOKEN_OFFSET) % 4096, |
|
|
(slots[6] - CODE_TOKEN_OFFSET) % 4096, |
|
|
]) |
|
|
|
|
|
return [l1, l2, l3] |
|
|
|
|
|
@torch.inference_mode() |
|
|
def decode( |
|
|
self, |
|
|
snac_tokens: List[int], |
|
|
use_sliding_window: bool = False |
|
|
) -> Optional[np.ndarray]: |
|
|
""" |
|
|
Decode SNAC tokens to audio waveform. |
|
|
|
|
|
Args: |
|
|
snac_tokens: List of SNAC token IDs (7*n tokens) |
|
|
use_sliding_window: If True, return only middle 2048 samples |
|
|
(for smooth streaming without pops/clicks) |
|
|
|
|
|
Returns: |
|
|
Audio waveform as float32 numpy array, 24kHz mono |
|
|
""" |
|
|
if len(snac_tokens) < SNAC_TOKENS_PER_FRAME: |
|
|
return None |
|
|
|
|
|
|
|
|
levels = self.unpack_snac_from_7(snac_tokens) |
|
|
|
|
|
if not levels[0]: |
|
|
return None |
|
|
|
|
|
|
|
|
codes = [ |
|
|
torch.tensor(level, dtype=torch.long, device=self.device).unsqueeze(0) |
|
|
for level in levels |
|
|
] |
|
|
|
|
|
|
|
|
z_q = self.snac_model.quantizer.from_codes(codes) |
|
|
audio = self.snac_model.decoder(z_q) |
|
|
|
|
|
|
|
|
audio = audio[0, 0].cpu().numpy() |
|
|
|
|
|
|
|
|
|
|
|
if use_sliding_window and len(audio) >= 4096: |
|
|
audio = audio[2048:4096] |
|
|
|
|
|
return audio |
|
|
|
|
|
def decode_to_bytes( |
|
|
self, |
|
|
snac_tokens: List[int], |
|
|
use_sliding_window: bool = False |
|
|
) -> Optional[bytes]: |
|
|
""" |
|
|
Decode SNAC tokens to audio bytes (int16 PCM). |
|
|
|
|
|
Args: |
|
|
snac_tokens: List of SNAC token IDs |
|
|
use_sliding_window: Use sliding window for smooth streaming |
|
|
|
|
|
Returns: |
|
|
Audio as bytes (int16 PCM, 24kHz mono) |
|
|
""" |
|
|
audio = self.decode(snac_tokens, use_sliding_window=use_sliding_window) |
|
|
|
|
|
if audio is None: |
|
|
return None |
|
|
|
|
|
|
|
|
audio_int16 = (audio * 32767).astype(np.int16) |
|
|
return audio_int16.tobytes() |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
class OnlyAudioAfterSOS: |
|
|
""" |
|
|
Restricts vocabulary to SNAC codes + EOS after SOS token. |
|
|
|
|
|
This prevents the model from generating text tokens during audio phase, |
|
|
which would cause "hallucination" where the model repeats description text |
|
|
instead of generating proper audio codes. |
|
|
""" |
|
|
|
|
|
def __init__(self): |
|
|
self._seen_sos = False |
|
|
|
|
|
def __call__( |
|
|
self, |
|
|
prompt_token_ids: List[int], |
|
|
generated_token_ids: List[int], |
|
|
logits: torch.Tensor, |
|
|
) -> torch.Tensor: |
|
|
""" |
|
|
Apply constraint: after SOS, only allow SNAC codes + EOS. |
|
|
|
|
|
Args: |
|
|
prompt_token_ids: Original prompt token IDs |
|
|
generated_token_ids: Tokens generated so far |
|
|
logits: Logits for next token [vocab_size] |
|
|
|
|
|
Returns: |
|
|
Modified logits with masked tokens |
|
|
""" |
|
|
|
|
|
if not self._seen_sos: |
|
|
all_token_ids = prompt_token_ids + generated_token_ids |
|
|
if CODE_START_TOKEN_ID in all_token_ids: |
|
|
self._seen_sos = True |
|
|
else: |
|
|
return logits |
|
|
|
|
|
|
|
|
mask = torch.full_like(logits, float('-inf')) |
|
|
mask[SNAC_MIN_ID:SNAC_MAX_ID + 1] = 0 |
|
|
mask[CODE_END_TOKEN_ID] = 0 |
|
|
|
|
|
return logits + mask |
|
|
|
|
|
def reset(self): |
|
|
"""Reset state for reuse across generations.""" |
|
|
self._seen_sos = False |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
class Maya1VoiceModel: |
|
|
""" |
|
|
Maya-1-Voice TTS Model with VLLM inference engine. |
|
|
|
|
|
Handles model loading, tokenizer initialization, and VLLM engine setup. |
|
|
""" |
|
|
|
|
|
def __init__( |
|
|
self, |
|
|
model_path: str, |
|
|
dtype: str = "bfloat16", |
|
|
max_model_len: int = 8192, |
|
|
gpu_memory_utilization: float = 0.85, |
|
|
): |
|
|
""" |
|
|
Initialize Maya-1-Voice model with VLLM. |
|
|
|
|
|
Args: |
|
|
model_path: Path to model checkpoint (local or HuggingFace) |
|
|
dtype: Model precision (bfloat16 recommended) |
|
|
max_model_len: Maximum sequence length |
|
|
gpu_memory_utilization: GPU memory fraction to use (0.0-1.0) |
|
|
""" |
|
|
self.model_path = model_path |
|
|
|
|
|
print(f"๐ Initializing Maya-1-Voice Model") |
|
|
print(f"๐ Model: {model_path}") |
|
|
print(f"๐ข Dtype: {dtype}") |
|
|
|
|
|
|
|
|
print(f"๐ Loading tokenizer...") |
|
|
self.tokenizer = AutoTokenizer.from_pretrained( |
|
|
model_path, |
|
|
trust_remote_code=True, |
|
|
) |
|
|
print(f"โ
Tokenizer loaded: {len(self.tokenizer)} tokens") |
|
|
|
|
|
|
|
|
print(f"๐ง Initializing VLLM engine...") |
|
|
engine_args = AsyncEngineArgs( |
|
|
model=model_path, |
|
|
tokenizer=model_path, |
|
|
dtype=dtype, |
|
|
max_model_len=max_model_len, |
|
|
gpu_memory_utilization=gpu_memory_utilization, |
|
|
trust_remote_code=True, |
|
|
) |
|
|
|
|
|
self.engine = AsyncLLMEngine.from_engine_args(engine_args) |
|
|
print(f"โ
VLLM engine ready") |
|
|
|
|
|
def build_prompt(self, description: str, text: str) -> str: |
|
|
""" |
|
|
Build prompt in Maya-1-Voice format using chat template. |
|
|
|
|
|
Format: Chat template with <description="..."> text as content |
|
|
|
|
|
The model expects: |
|
|
1. Description of voice/character |
|
|
2. Text to synthesize (optionally with <emotion> tags) |
|
|
|
|
|
Args: |
|
|
description: Voice description |
|
|
Example: "Realistic male voice in the 30s age with american accent. |
|
|
Normal pitch, warm timbre, conversational pacing." |
|
|
text: Text to synthesize |
|
|
Example: "Hello world! <excited> This is amazing!" |
|
|
|
|
|
Returns: |
|
|
Formatted prompt string using chat template |
|
|
""" |
|
|
content = f'<description="{description}"> {text}' |
|
|
messages = [{"role": "user", "content": content}] |
|
|
return self.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
class Maya1VoiceStreamingPipeline: |
|
|
""" |
|
|
Streaming TTS pipeline using sliding window approach. |
|
|
|
|
|
This generates smooth audio by: |
|
|
1. Streaming tokens from VLLM as they're generated |
|
|
2. Every 7 tokens, decoding the last 28 tokens (4 frames) - sliding window |
|
|
3. Keeping only middle 2048 samples from each decode |
|
|
4. Creating natural overlap between chunks for artifact-free playback |
|
|
""" |
|
|
|
|
|
def __init__(self, model: Maya1VoiceModel, snac_decoder: SNACDecoder): |
|
|
"""Initialize streaming pipeline.""" |
|
|
self.model = model |
|
|
self.snac_decoder = snac_decoder |
|
|
print(f"๐ Maya-1-Voice Streaming Pipeline initialized") |
|
|
|
|
|
async def generate_speech_stream( |
|
|
self, |
|
|
description: str, |
|
|
text: str, |
|
|
temperature: float = DEFAULT_TEMPERATURE, |
|
|
top_p: float = DEFAULT_TOP_P, |
|
|
max_tokens: int = DEFAULT_MAX_TOKENS, |
|
|
repetition_penalty: float = DEFAULT_REPETITION_PENALTY, |
|
|
) -> AsyncGenerator[bytes, None]: |
|
|
""" |
|
|
Generate speech audio with streaming. |
|
|
|
|
|
Args: |
|
|
description: Voice/character description |
|
|
text: Text to synthesize (with optional <emotion> tags) |
|
|
temperature: Sampling temperature (lower = more stable) |
|
|
top_p: Nucleus sampling |
|
|
max_tokens: Max SNAC tokens to generate |
|
|
repetition_penalty: Prevent repetition loops |
|
|
|
|
|
Yields: |
|
|
Audio chunks as bytes (int16 PCM, 24kHz mono) |
|
|
""" |
|
|
print(f"\n๐ Starting streaming generation") |
|
|
print(f"๐ Description: {description[:80]}...") |
|
|
print(f"๐ฌ Text: {text}") |
|
|
|
|
|
|
|
|
prompt = self.model.build_prompt(description, text) |
|
|
|
|
|
|
|
|
sampling_params = SamplingParams( |
|
|
temperature=temperature, |
|
|
top_p=top_p, |
|
|
max_tokens=max_tokens, |
|
|
min_tokens=DEFAULT_MIN_TOKENS, |
|
|
repetition_penalty=repetition_penalty, |
|
|
stop_token_ids=[CODE_END_TOKEN_ID], |
|
|
) |
|
|
|
|
|
print(f"๐ฒ Sampling: temp={temperature}, top_p={top_p}, max_tokens={max_tokens}") |
|
|
|
|
|
|
|
|
token_buffer = [] |
|
|
total_tokens = 0 |
|
|
total_chunks = 0 |
|
|
|
|
|
|
|
|
import uuid |
|
|
import time |
|
|
request_id = f"maya1voice-{uuid.uuid4().hex[:8]}-{int(time.time() * 1000000)}" |
|
|
|
|
|
results_generator = self.model.engine.generate( |
|
|
prompt=prompt, |
|
|
sampling_params=sampling_params, |
|
|
request_id=request_id, |
|
|
) |
|
|
|
|
|
|
|
|
async for request_output in results_generator: |
|
|
generated_ids = request_output.outputs[0].token_ids |
|
|
|
|
|
|
|
|
new_tokens = generated_ids[total_tokens:] |
|
|
total_tokens = len(generated_ids) |
|
|
|
|
|
|
|
|
for token_id in new_tokens: |
|
|
if SNAC_MIN_ID <= token_id <= SNAC_MAX_ID: |
|
|
token_buffer.append(token_id) |
|
|
|
|
|
|
|
|
|
|
|
if len(token_buffer) % 7 == 0 and len(token_buffer) > 27: |
|
|
window_tokens = token_buffer[-28:] |
|
|
|
|
|
|
|
|
audio_bytes = self.snac_decoder.decode_to_bytes( |
|
|
window_tokens, |
|
|
use_sliding_window=True |
|
|
) |
|
|
|
|
|
if audio_bytes: |
|
|
total_chunks += 1 |
|
|
if total_chunks == 1: |
|
|
print(f"๐ต First chunk decoded ({len(audio_bytes)} bytes)") |
|
|
yield audio_bytes |
|
|
|
|
|
print(f"โ
Streaming complete: {total_tokens} tokens โ {total_chunks} chunks") |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
async def main(): |
|
|
""" |
|
|
Example usage of Maya-1-Voice streaming inference. |
|
|
|
|
|
This demonstrates: |
|
|
1. Model initialization |
|
|
2. SNAC decoder setup |
|
|
3. Streaming generation |
|
|
4. Audio chunk handling |
|
|
""" |
|
|
|
|
|
|
|
|
MODEL_PATH = "/home/ubuntu/veena_temp/maya-1-voice" |
|
|
DEVICE = "cuda" if torch.cuda.is_available() else "cpu" |
|
|
|
|
|
print("=" * 80) |
|
|
print("Maya-1-Voice VLLM Streaming Inference Example") |
|
|
print("=" * 80) |
|
|
|
|
|
|
|
|
model = Maya1VoiceModel( |
|
|
model_path=MODEL_PATH, |
|
|
dtype="bfloat16", |
|
|
max_model_len=8192, |
|
|
gpu_memory_utilization=0.8, |
|
|
) |
|
|
|
|
|
|
|
|
snac_decoder = SNACDecoder(device=DEVICE) |
|
|
|
|
|
|
|
|
pipeline = Maya1VoiceStreamingPipeline(model, snac_decoder) |
|
|
|
|
|
|
|
|
description = ( |
|
|
"Realistic male voice in the 30s age with american accent. " |
|
|
"Normal pitch, warm timbre, conversational pacing, neutral tone delivery at med intensity." |
|
|
) |
|
|
text = "Hello! This is a test of the Maya-1-Voice text-to-speech system." |
|
|
|
|
|
print(f"\n{'='*80}") |
|
|
print("Example 1: Professional Voice") |
|
|
print(f"{'='*80}") |
|
|
|
|
|
audio_chunks = [] |
|
|
async for chunk in pipeline.generate_speech_stream( |
|
|
description=description, |
|
|
text=text, |
|
|
temperature=0.4, |
|
|
max_tokens=500, |
|
|
): |
|
|
audio_chunks.append(chunk) |
|
|
print(f"๐ฆ Received chunk {len(audio_chunks)}: {len(chunk)} bytes") |
|
|
|
|
|
|
|
|
full_audio = b''.join(audio_chunks) |
|
|
print(f"\nโ
Total audio: {len(full_audio)} bytes ({len(full_audio)//2} samples, {len(full_audio)/2/24000:.2f}s)") |
|
|
|
|
|
|
|
|
try: |
|
|
import wave |
|
|
output_file = "output_example1.wav" |
|
|
with wave.open(output_file, 'wb') as wav: |
|
|
wav.setnchannels(1) |
|
|
wav.setsampwidth(2) |
|
|
wav.setframerate(24000) |
|
|
wav.writeframes(full_audio) |
|
|
print(f"๐พ Saved to {output_file}") |
|
|
except ImportError: |
|
|
print(f"โ ๏ธ Install 'wave' module to save audio files") |
|
|
|
|
|
|
|
|
print(f"\n{'='*80}") |
|
|
print("Example 2: Character Voice with Emotions") |
|
|
print(f"{'='*80}") |
|
|
|
|
|
description = ( |
|
|
"Creative, dark_villain character. Male voice in their 40s with british accent. " |
|
|
"Low pitch, gravelly timbre, slow pacing, angry tone at high intensity." |
|
|
) |
|
|
text = "The darkness isn't coming... <angry> it's already here!" |
|
|
|
|
|
audio_chunks = [] |
|
|
async for chunk in pipeline.generate_speech_stream( |
|
|
description=description, |
|
|
text=text, |
|
|
temperature=0.5, |
|
|
max_tokens=800, |
|
|
): |
|
|
audio_chunks.append(chunk) |
|
|
print(f"๐ฆ Received chunk {len(audio_chunks)}: {len(chunk)} bytes") |
|
|
|
|
|
full_audio = b''.join(audio_chunks) |
|
|
print(f"\nโ
Total audio: {len(full_audio)} bytes ({len(full_audio)//2} samples, {len(full_audio)/2/24000:.2f}s)") |
|
|
|
|
|
|
|
|
try: |
|
|
import wave |
|
|
output_file = "output_example2.wav" |
|
|
with wave.open(output_file, 'wb') as wav: |
|
|
wav.setnchannels(1) |
|
|
wav.setsampwidth(2) |
|
|
wav.setframerate(24000) |
|
|
wav.writeframes(full_audio) |
|
|
print(f"๐พ Saved to {output_file}") |
|
|
except ImportError: |
|
|
pass |
|
|
|
|
|
print(f"\n{'='*80}") |
|
|
print("๐ Examples complete!") |
|
|
print(f"{'='*80}") |
|
|
|
|
|
|
|
|
if __name__ == "__main__": |
|
|
|
|
|
asyncio.run(main()) |